Art in the Age of Misanthropic Reproduction

Noah Dillon

January 29, 2018

I was working for a painter and one day an art dealer emailed the studio explaining that they were considering purchasing, on the secondary market, a supposed preparatory drawing made in the 1980s by my boss. The dealer wanted an opinion on its legitimacy before buying this expensive thing from a third party. The drawing resembled paintings my boss had made, but only notionally: the colors were all wrong, as were the materials, and the execution. The back of the frame had various misspelled labels intended to demonstrate provenance. On the verso, recapitulating the artwork’s claim to veracity, the clumsy forger had invented a signature for my boss and had jotted above it “this is my work.”¹

This sort of thing is not uncommon. Georgina Adam, author of The Dark Side of the Boom (2018), posits that a lot of forgery cases are for prints or lesser artists—that the market for such bunkum is substantial. Anthony M. Amore, in The Art of the Con (2016) cites estimates of 40–50% of all artworks in museums or at market. Roman workshops created fake Greek statuary. Michelangelo, in his day, created ersatz Roman sculptures. Albrecht Dürer, exasperated by bootleg copies of his work, included in an edition of his prints a threat of legal sanction and corporal mutilation for would-be forgers. In addition to seizing actual invaluable works, Hermann Goering was also duped by a clever faker. A recent article in the New York Times noted the rising problem of fake artworks in Vietnam’s museums and galleries, arising from gaps in provenance caused by a multi-decade imperialist war. (A similar problem arising from World War II was a boon to panderers of fakes in the decades that followed that conflict.)

Orson Welles, of course, famously chronicled the forger Elmyr de Hory, and his biographer, Clifford Irving, who was also a forger. Ely Sakhai commissioned and sold forged works by Marc Chagall, Paul Gauguin and Pierre-Auguste Renoir. He was caught in the early 2000s when he tried to sell the originals at the same time collectors tried to resell the fakes he’d sold them. In 2010, Christian Parisot, a highly regarded expert and historian of Amedeo Modigliani—President of the Modigliani Institute for Christ’s sake—was convicted of peddling forged drawings attributed to Jeanne Hébuterne, Modigliani’s mistress. Dealer John Natsoulas attributed Henry Villierme’s work to Richard Diebenkorn.

More deliciously, there are the infamous fake Abstract Expressionist paintings by Pei Shen-Qian, which, when discovered, led to the literal overnight closing of the prestigious, 165-year-old gallery M. Knoedler & Co. Dealers, critics, collectors, scholars, other reputable artists had given at least minimal thumbs-up to Qian’s work, and they now faced scorn (and in some cases, legal action) for their poor judgment and lack of due diligence—for fucking up.²

ABC’s Nightline interviewed Qian in Shanghai, where he lives and avoids extradition. He claimed that he didn’t intend for his paintings to be mistaken for original artworks and sold for millions of dollars. Rather, they were novelties. They were homages, decorations, “just copies,” not forgeries; their resemblance was barely more than incidental, basically, according to Qian. He told the FBI that the whole thing was “a very big misunderstanding.” (And plus, he says, he didn’t get compensated adequately for his crucial role in the alleged crime, which he claims he didn’t commit.)

And there are less sinister doppelgangers: the copying and one-upmanship of Pablo Picasso and Georges Braque, or the innumerable painters of monochromes and grids, of splattery or painterly abstractions, redundant Medieval Biblical scenes attributed to this artist’s workshop and then that one.³ The value of art is commonly believed to reside in the authentic creations of unique humans, subject to verification by thoughtful viewers; but it very clearly does not. Were that the case, much of art history would instantly combust.

There will be more imposters, whether confidence men or redundancies. It is inevitable that some of them will be machines, either by accident or by design. Machines will soon think and some of them will make things that could be art. (Is art made by robots synonymous with art that is fake or redundant? Is that the suggestion here?) Forgeries, machine-made artworks, art-like detritus—these are border cases: products that look like art, but are not unimpeachably art.

Those margins have grown repeatedly. Last year was the centenary of Marcel Duchamp’s “Fountain,” a sculpture fabricated by a large, mechanized process with no artistic intent at all prior to his intercession.⁴ It was followed by the Futurists’ predictions of—and demands for—humanity’s replacement by machines. Jean Tinguely’s kinetic sculpture, “Homage to New York” (1960), a big machine intended to run itself into death, was unleashed at MoMA, briefly sustained functionality as an absurd contraption, then proceeded into its own terminus as a flaming hulk that didn’t even self-destruct properly.

Gustav Metzger produced works that were likewise completed with their destruction; elsewhere, he confronted viewers with images of our own self-annihilation, such as in war and in mass movements. Andy Warhol coolly remarked in a 1963 interview that, “I feel that whatever I do and do machine-like is what I want to do.” Vito Acconci is supposed to have said, “The artist must be a kind of dumb copying machine.” Jasper Johns, in a 1964 notebook, jotted down an algorithm for artmaking: “Take an object / Do something to it / Do something else to it. [Repeat.]”

Roxy Paine rushed toward this future in the late 1990s, with his painting and sculpting machines pumping out surprising, sludgy, horrid and monstrous monochromatic things... Caleb Larsen’s “A Tool to Deceive and Slaughter” (2009) is a box that is only activated as an artwork when it’s hooked to an Ethernet connection, which it uses to automatically list itself for sale on eBay. Larsen’s sculpture is only art if it’s being sold, and a collector only owns the machine until it sells itself to another collector.

For the past two decades, Siebren Versteeg has written algorithms that produce images for him to print or present as videos, drawing from or echoing an increasing number of easily produced Markov bots which randomly compile snippets of text to form a kind of absurd poetry on Twitter and elsewhere—every bit as aleatory as Duchamp himself. Kenneth Goldsmith and Christian Bök, the former a skeptic, the latter a propagator, have already started addressing these problems, predicting a future of limitless robot speech.⁵

Automatically generated material is proliferating, but these are only tremors in what is likely to be even broader disruption. We are not yet at the point of general purpose artificial intelligence, but it is not inconceivable that machine learning will soon be sophisticated and independent to such a degree that the connection between programmer(s) and a program’s output is indirect enough to be the equivalent of the influence of a parent over the actions of their adult child.⁶ Our beliefs and expectations about what a machine art can be are probably underdeveloped and wrong (including this very speculative essay itself).

My brother, interested in artificial intelligence and machine learning, sends me demonstrations of breakthroughs in computer science and data crunching. Some clever team has made an incremental improvement on the work of their predecessors and can more and more ask their purpose-built machines to produce an image of a bird or a flower or another noun, and the machine will invent—new and original—an image of the requested thing. Off the top of its head, so to speak.

A friend asks, about such robots and self-driving cars, “Where does this road end?”⁷ A world is coming in which stuff—intellectual and physical—will be produced by machines or bots or autos or whatever we call them. Such production will have no human input, and some of said stuff could be recognized as art—stuff for its own sake, not as a byproduct or means to an end; stuff that, if it had been made by humans called artists, would have been called art. In our future we face the prospect of artworks made by no human and for no human.

Automation is decimating labor, as well as upending long-held beliefs and assumptions about work.⁸ Machines increasingly replace people in the workforce, from supermarket checkout lanes to assembly lines to law offices. Some scholars, such as Jeremy Howard, have predicted that nearly 50% of existing jobs will likely be automated in the near future.⁹ Machines are typically more productive, more energy efficient, more accurate than humans. They do not get sick, don’t get tired, are not prejudiced, and you can’t insult them.¹⁰ They don’t suffer from the anxiety of influence.

And this is not inherently bad. People should not have to do a lot of the work they currently do. Repetitive, low-skill, degrading labor often harms people. George Orwell, in Down and Out in Paris and London (1933), notes that “we have made a sort of fetish of manual work,” glorifying it as noble impoverishment that abuses and confines people. The low value of such work grinds a person’s body and mind into dust. It’s frequently demeaning and discriminatory. Neither consumers nor management actually appreciate that labor, and so it pays poorly with few or no benefits or security. It requires long hours, inconsistently parceled out. It can be very resource intensive, including the calories and carbon consumed by workers. 

But automation is unlikely to stop with factories or low-skill jobs. TV shows like Star Trek predict a future in which even cooks are missing, replaced by dispensers or pills, and medical attention is administered by electronic assistants. No one in those realms appears bothered about this. On the chopping block here in reality, probably sooner than we imagine, are all drivers, whole blocs of manual laborers, paralegals, many of the few remaining journalists, diagnosticians, retailers, analysts of all kinds: economists and brokers and supply chain managers and medical researchers. And the list goes on.

Even if those predictions are exaggerated by an order of magnitude (unlikely), that’s still a very large class of jobless people. Unfortunately, new machines, capable of thinking and working in inhuman ways, also come now, at a time of low power among laborers. The union movement has been crushed to near death, or at least dormancy, in the United States.¹¹ The checks against capitalism’s apparently innate process of funneling wealth upwards, the imposition of mechanisms for redistributing wealth from the capitalist class to the working class—which were wimpy to begin with—are ever more ash in the wind.

And here are artists, who, like drivers and burger-flippers and riveters and cetera, stand in lousy relation to networks of labor and capital from the outset. There’s little history of unionization; the economic model now in place (better than the pre-modern era, certainly, but still awful) disenfranchises most artists and has few protections against the rapaciousness of those individuals or institutions with money or power.

In mainstream culture, the pedestrian idea of what constitutes an artist appears basically stuck in Romantic 19th-century aristocratic norms: a white male, impoverished but mysteriously economically sustainable (a trust fund or benefactor maybe? Eating on credit, or from a dumpster?); with a bohemian lifestyle, empty head, full heart, and naïvely, selfishly libertine social values.¹² The artist is idiosyncratic and sui generis. They’re depicted as dilettantes who work to express amorous fancies that are more important to them than making a living.¹³ Artists, being considered “special people,” are seen as outside of society, and so too its legal checks on power and threadbare social safety nets.

(And all this despite an ambivalent cultural lip service to the idea that great art is dependent on great wealth and that distinguishing “good taste” in art is essential to good citizenship and human well-being.)¹⁴

One thing that should be noted is that since Duchamp, at least, the role of the artist has gotten kind of confusing and expansive, as those earlier examples of near-mechanized art attest. Entirely novel modes of production and reproduction have been incorporated into the arts, as have new ways of thinking and talking about all this stuff, a new syntax and vocabulary and even values. And, excitingly, the membrane between what art is and is not has become super porous.

So then automation might be a good test case to consider how we relate to the arts and to human labor. What, if anything, about human productivity is too invaluable to leave to machines or dispense with entirely, and why? And what, if anything, about art is essential and useful and good, and why? And further, to consider the world in which we aren’t the only sentients, or after we are extinct: what are we to make of a world that is inhuman, and/or in which art is inhuman?¹⁵ The philosopher Peter Singer, in his book Animal Liberation, kicked the door to such considerations wide open.

This isn’t merely about machines overtaking people. Instead, the question is how do we now account for artworks—such as those found on the city street or in cyberspace—without a human artist, a person as author? How do we imagine a world beyond us and even without us?

It could be that these questions and their answers are completely untroubling to everyone. Art is strictly a human endeavor—probably, for now, at least on this planet. Walter Benjamin, in his famous essay on reproduced images, talks about the toll mechanization takes on the artwork and speculates about the assumed relation to the viewer, but doesn’t address an image’s economics of power: seduction. If there is some causal and beneficial effect for machines producing likeable images, you can bet automata will figure it out, eventually.

Algorithms are very well equipped to figure out what consumers want, and to produce and serve us such content, even merely through the coercion of, say, a frowny Mac icon begging you to tend to your machine. A clever bot can displace people—including artists. It could be that most people just want, regardless of authorship, images they like, and images that have already been validated by the culture, because they are similar to or are identical with existing pictures, or they’re simply really cool. One can imagine a “Warhol” made by an artificially intelligent machine, because people like the image that the machine gives them, and the machine likes that humans like the image, for whatever reason.

Futurism predicted a machine art that would be dynamic and spectacular, when in fact from today’s vantage point it seems more and more that machinic art will probably be super dull. It appears unlikely that people will find very interesting any artworks produced by machines. How can a robot say anything meaningful to humans?

Nevertheless, such art will be made, and it could be titillating. 

(And then there is that question asked of art, that is not currently asked of machine-made products: whether the work is “original,” meaning that the question regresses back to the maker, not the object.)

Appreciating fully the demise of novelty’s privileged place in the arts since Modernism, and understanding more fully that most of the plastic arts’ wheels have been reinvented over and over throughout time, might artists be freed by a confrontation with a black box that produces an image because its value has already been demonstrated? “Take pity on our poor authors!” writes McKenzie Wark, in a teasing passage of The Beach Beneath the Street (2011) about the plagiarism scandals of various historians. “Not even they can tell their own words from another’s. They are caught between the monotonous consistency of official historical narratives and the demand that the middle-class author have a unique vision that is his or her personal property.” Perhaps that pressure could vanish.

(Are forgeries simply more artworks? Is that the implication here?)

This essay developed from photographs of contingent, incidental forgeries: things that are not artworks, that are not made with an artist's intent, and nonetheless look like artworks, especially those by famous artists. I initially wanted to know what the point of making art was when the world is already pervaded by art-like things made acceptable (formally, materially, conceptually) by the history of 20th-century arts.¹⁶ People are pareidolic, and they find art where they look.¹⁷

What, in practice, separates John Ruskin’s “pathetic fallacy”—the poet’s anthropomorphization of things in the world—and a successful subject in a Turing test? In each case, humans cede portions of a quality previously believed to be essential and ours alone, perhaps best characterized as sentience. Here comes the encroachment of machines into realms formerly reserved only for sentients, for people. Such realms have, in fact, helped define what it is to be human, which seems, at least this far out, way harder to come to grips with than new inventions, changing norms around screens, the displacement of bank clerks by ATMs, etc. The encroachment of non-human beings could help define sentience, humanity, art, work, and so on, by showing us how much larger than ourselves these things are.

This began with Duchamp. And the apocalyptic tone of machine takeover, not as Terminator-type war but as slow seep, is making (or will make?) people more redundant. Imagine Terminator’s most potent image, of an anthropomorphic robot crushing a human skull with its foot, only instead of a machine gun, the robot holds a brush in its articulated, animatronic hand, painting a memento mori. In Duchamp’s demonstration of art’s accessibility to anyone willing to declare an object as their art and themselves as artists, it feels kind of inevitable that we’d have to share that distinction with other beings, whether the existing organic ones or the rising algorithmic ones or anything else.

1. This idiotic ploy, aimed at an even more idiotic skeptic, was funny: if you doubt that this work is real, have no fear: the artist has scrawled an assurance there for you, for just this kind of dubious situation.
    
2. At least a few of these people, such as David Anfam and Jack Flam, still hold some modicum of respect among art world managers, though why is completely bewildering. Take Anfam, who wrote Mark Rothko’s catalogue raisonne: Artnet News reports that he validated a theretofore unknown Rothko painting as authentic without examining it and without interrogating its bullshitty provenance. It was totally fake. He lobbied the Albright-Knox to purchase a fake Barnett Newman painting from Knoedler and negotiated a finder's fee for himself for the sale of that high-price bunkum. He was further accused of hearing out a proposal from the Knoedler scammers to endorse the attribution of a faked scrap of a supposedly lost Clyfford Still painting to the Clyfford Still Museum, where he is employed as Senior Consulting Curator. He had included faked artworks in exhibitions he’d curated.
   
   And this shady nonsense continued even after rumors and allegations had started being made against Anfam and Knoedler. There was a lot of lying and willful suspension of critical thinking in the Knoedler case, among all parties, as in all such cases, so Anfam was likely duped as much as he was actively participating in these charades. The dealer at the center of this storm bought a fake painting attributed to Jackson Pollock that was signed, ineptly, “Pollok.” They and others bought into and participated in blatant, stupid fraud. Despite this, several months after his embarrassing admissions about all this in US District Court, a show of AbEx art curated by Anfam opened in London at the Royal Academy of Arts, and later traveled to the Guggenheim Bilbao, and the (almost exclusively) men included in the show are also some of those whose faked paintings were explicitly promoted by Anfam, for profit, which should make one ask: 1. Why is David Anfam considered a reliable source of ideas and information about any art, and especially AbEx painting? and 2. Even given the expense and time and difficulty in putting together a museum exhibition, why would institutions and lenders stand by a known abettor to and panderer of forgeries?
    
3. Is a $450-million portrait of Christ actually by Leonardo, or his apprentices, or not? Is the identity of the work, like the price, determined by what the market can bear?
    
4. N.b. that Duchamp’s “Bicycle Wheel” (1913) was actually the first readymade he created, several years before “Fountain”, which was his first readymade to be exhibited. Note also the simultaneous centennials of the readymade and Suprematism, the end-game art movement Kazimir Malevich claimed to have founded in 1913. In the mid-1990s, Mike Bidlo made imitations of both “Fountain” and “Bicycle Wheel”—a kind of cheap move that nonetheless puts an interesting question mark on the site of authorship in these objects.
    
5. It’s worth quoting Goldsmith on this at length:
   
   Christian Bök calls this robopoetics, a condition where “the involvement of an author in the production of literature has henceforth become discretionary.” He asks, “why hire a poet to write a poem when the poem can in fact write itself?” Science fiction is poised to become reality, enacting Bök’s prediction for the literary future when he claims that, “we are probably the first generation of poets who can reasonably expect to write literature for a machinic audience of artificially intellectual peers. Is it not already evident by our presence at conferences on digital poetics that the poets of tomorrow are likely to resemble programmers, exalted, not because they can write great poems, but because they can build a small drone out of words to write great poems for us? If poetry already lacks any meaningful readership among our own anthropoid population, what have we to lose by writing poetry for a robotic culture that must inevitably succeed our own? If we want to commit an act of poetic innovation in an era of formal exhaustion, we may have to consider this heretofore unimagined, but nevertheless prohibited, option: writing poetry for inhuman readers, who do not yet exist, because such aliens, clones, or robots have not yet evolved to read it.”
   But this begs the question does literature need to be read? Or more pressingly, is it actually read?
   
   It would seem that perhaps this essay itself is redundant.
    
6. General-purpose AI is a type of machine that can learn on its own, unmediated, and can develop an aptitude for essentially any task. What might be interesting here is that such machines could more easily than humans fulfill Karl Marx’s dream, in “The German Ideology,” of “communist society, where nobody has one exclusive sphere of activity but each can become accomplished in any branch he wishes, society regulates the general production and thus makes it possible for me to do one thing today and another tomorrow, to hunt in the morning, fish in the afternoon, rear cattle in the evening, criticize after dinner, just as I have a mind, without ever becoming hunter, fisherman, herdsman or critic.” Maybe because machines start from a position outside of class…?
    
7.  Heh.
    
8. For example, contrary to popular belief, the United States makes more goods now than ever before and that productivity is rising. It is the most productive nation in the world most years. But machines do much of that labor, with fewer human technicians of higher skill and educational levels overseeing those robotized processes. And maybe even those jobs will be exterminated someday.
   
   And as even as I write this, the New York Times has just published an essay by Thomas B. Edsall, with a really persuasive analysis of overlapping phenomena of replacement by robots in white, male, low-skill labor, and that cohort’s political revanchism and support for Donald Trump.
    
9. cf. CGP Grey’s video “Humans Need Not Apply” and a special report on “Automation and anxiety” in a June 2016 issue of the Economist. Also Jeremy Howard’s discussion of the “The Data Science Revolution” and the dismal or utopian possibilities of machine learning that we have the opportunity, and obligation, to make choices about right away.
    
10. A promotional video for Boston Dynamics’ humanoid robot Atlas features the dumb thing being basically assaulted by a developer as the robot tries to carry boxes around, in order to demonstrate its resilience and adaptability, its ability to continue working in spite of disruptions. But the metallic hulk seems anthropomorphically sad, lonely, pathetic, rouses sympathy, and the video could be re-imagined as a futuristic primer on workplace bullying. These are, of course, anthropic projections on emotionless steel and silicon.
     
11. Union membership sank lower than ever in 2016, according to reporting by NPR and BuzzFeed. And please Nb: they report that this decline isn’t merely confined to manufacturing (whoever said only manufacturing can be unionized?), but includes, among other things, public sector employees, too, where upwards of 70% of workers had formerly been unionized, as in New York.
     
12. And doesn’t much of this stereotype extend to other low-wage and low-skill laborers? Yes.
     
13. e.g. to quote A. O. Scott: “The way we habitually think and talk about [the relationship between art and money] betrays a deep and venerable ambivalence. On one hand, art is imagined to exist in a realm of value that lies beyond and beneath mere economic considerations. The old phrase ‘starving artist’ gestures toward an image that is both romantic and pathetic, of a person too pure, and also just too impractical, to make it in the world. When that person ceases to starve, he or she can always be labeled a sellout. You’re not supposed to be in it for the money.”
     
14. A friend notes that Artist, like a lot of professions, is totally misunderstood by those outside of its market.
     
15. I’m just going to say, for the record, that I don’t consider human extinction, in the grand scheme of things, as such a catastrophe. It’s going to happen, eventually, somehow. The suffering of any organism is a tragedy that should be prevented. But we will someday be gone and the planet will continue and I think that’s OK. We do, however, have the ability (and therefore kind of the obligation, ethically speaking) to consider what kind of world we leave in the wake of our extinction.
    
    Cue the Butthole Surfers’ “One Hundred Million People Dead” or Cows’ “Peacetika.”
     
16. Even this essay is preceded by—inadvertently plagiarizes—Ernst Carmella's “Forty Found Fakes, 1978–1979” (1979), which is an almost identical project of artists and their accidental cultural antecedents.
    
    This essay, this footnote, is indeed redundant.
     
17. A friend offers this rejoinder: “[Regardless] of its origins, the results do not become art until a human interprets the object as art. Art that is never viewed by a human thinking ‘I am looking at a piece of art’ is like the tree that falls in the forest. Regardless if the object is human-made, machine-made, or randomly encountered on the street, it's the act of interpretation that makes the art.”
    
    Would viewers reject art objects if they knew they had non-human origins? I’m not even certain the answer’s germane.